Glia cells contribute to motor neuron degeneration and aggravate the progression of amyotrophic lateral sclerosis (ALS), an idiopathic, fatal neurodegenerative disease of the human motor system. The nature of this contribution is not yet fully defined. Deficient expression and activity of the astroglia glutamate transporter EAAT2 (a.k.a. GLT-1 in rodents) has been reported in ALS and the excitotoxicity resulting from the ensuing accumulation of glutamate is one conventional mechanism that could contribute to motor neuron degeneration. We unraveled an unconventional way for EAAT2 to trigger motor neuron toxicity. We showed that caspase-3 selectively cleaves EAAT2 in ALS, presumably through a mechanism that involves restricted non-apoptotic caspase-3 activation in astrocytes, generating a cytosolic EAAT2 C-terminus SUMOylated fragment (CTE- SUMO1). This fragment, when exogenously expressed in astrocytes, accumulates in their nuclei in PML- nuclear bodies and indirectly causes motor neuron toxicity via a mechanism that involves increased expression and release of netrin-1. Oligodendrocytes, not astrocytes, secrete netrin-1 in the normal adult CNS to maintain axonal homeostasis. Hence, an abnormal non-physiological release of netrin-1 from CTE-SUMO1+ astrocytes could be one non-cell autonomous mechanism of toxicity to motor neurons in vivo as well. We propose in this application to expand the relevance of our in vitro findings by testing in vivo the hypothesis of whether CTE-SUMO1 is responsible for motor neuron impairment and ALS-like phenotype. To restate our goal in terms of specific aims, we propose: (1) To determine whether CTE-SUMO1 is a mediator of motor neuron impairment in vivo and assess the modalities of toxicity; (2) To determine whether preventing the accumulation of CTE-SUMO1 subsides the ALS phenotype. Aim 1 involves selective expression of CTE-SUMO1in ventral horn astrocytes of adult mice to evaluate its possible toxic effects on the entire phrenic motor neuron pool targeted via intraspinal focal injections of an AAV1-CTE-SUMO1 viral construct at the C4-C6 cervical level. Aim 2 involves the creation of a novel knock-in SOD1-G93A mouse model in which the caspase-3 cleavage site in EAAT2 is mutated to prevent the endogenous production of CTE-SUMO1. In addition to these in vivo experiments, we propose: (3) To determine the disease-relevant effectors responsible for CTE-SUMO1 accumulation in astrocytes. We expect to unravel the modalities of CTE-SUMO1 generation in astrocytes and whether the presence of familial ALS-linked causative mutations catalyzes the process. It is a matter of importance to understand what effectors could lead to the cleavage of EAAT2 and therefore the ensuing CTE-SUMO1 creation. This knowledge can then be applied to design pharmacological approaches aiming at arresting CTE-SUMO1 production in ALS. The scope of this last aim could then be key, in particular if results of aim 1-2 establisha role of the astroglial CTE-SUMO1 fragment in motor neuron toxicity and ALS pathogenesis.